| /external/tensorflow/tensorflow/contrib/optimizer_v2/ |
| D | rmsprop.py | 120 mom = state.get_slot(var, "momentum")
127 mom,
140 mom,
150 mom = state.get_slot(var, "momentum")
157 mom.handle,
168 mom.handle,
178 mom = state.get_slot(var, "momentum")
185 mom,
197 mom,
208 mom = state.get_slot(var, "momentum")
[all …]
|
| D | momentum.py | 81 mom = state.get_slot(var, "momentum")
84 mom,
92 mom = state.get_slot(var, "momentum")
95 mom.handle,
103 mom = state.get_slot(var, "momentum")
106 mom,
115 mom = state.get_slot(var, "momentum")
118 mom.handle,
|
| /external/guice/extensions/assistedinject/test/com/google/inject/assistedinject/ |
| D | ManyConstructorsTest.java | 275 assertEquals(null, pops.mom); in testDependenciesAndOtherAnnotations()
278 assertEquals("Mom", moms.mom); in testDependenciesAndOtherAnnotations()
281 assertEquals("Mom", momAndPop.mom); in testDependenciesAndOtherAnnotations()
287 Farm momsFarm(@Assisted("mom") String mom); in momsFarm() argument
289 Farm momAndPopsFarm(@Assisted("mom") String mom, @Assisted("pop") String pop); in momAndPopsFarm() argument
294 String mom; field in ManyConstructorsTest.Farm
302 Farm(@Assisted("mom") String mom, @Assisted("pop") String pop, Cow cow, Dog dog) { in Farm() argument
304 this.mom = mom; in Farm()
308 Farm(@Assisted("mom") String mom, Cow cow) { in Farm() argument
309 this.mom = mom; in Farm()
|
| /external/tensorflow/tensorflow/python/training/ |
| D | momentum.py | 96 mom = self.get_slot(var, "momentum")
98 var, mom,
106 mom = self.get_slot(var, "momentum")
108 var.handle, mom.handle,
116 mom = self.get_slot(var, "momentum")
118 var, mom,
126 mom = self.get_slot(var, "momentum")
128 var.handle, mom.handle,
|
| D | rmsprop.py | 142 mom = self.get_slot(var, "momentum")
149 mom,
160 mom,
170 mom = self.get_slot(var, "momentum")
177 mom.handle,
188 mom.handle,
198 mom = self.get_slot(var, "momentum")
205 mom,
217 mom,
228 mom = self.get_slot(var, "momentum")
[all …]
|
| /external/doclava/res/assets/templates/assets/ |
| D | doclava-developer-reference.js | 98 function new_node(me, mom, text, link, children_data, api_level) argument
103 node.depth = mom.depth + 1;
106 mom.get_children_ul().appendChild(node.li);
196 function get_node(me, mom) argument
198 mom.children_visited = true;
199 for (var i in mom.children_data) {
200 var node_data = mom.children_data[i];
201 mom.children[i] = new_node(me, mom, node_data[0], node_data[1],
287 var mom = me.node;
290 mom = mom.children[j];
[all …]
|
| /external/tensorflow/tensorflow/core/api_def/base_api/ |
| D | api_def_ResourceSparseApplyRMSProp.pbtxt | 16 name: "mom"
48 A vector of indices into the first dimension of var, ms and mom.
54 If `True`, updating of the var, ms, and mom tensors is protected
61 Note that in dense implementation of this algorithm, ms and mom will
63 and mom will not update in iterations during which the grad is zero.
69 mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)
70 var <- var - mom
|
| D | api_def_SparseApplyRMSProp.pbtxt | 16 name: "mom"
48 A vector of indices into the first dimension of var, ms and mom.
60 If `True`, updating of the var, ms, and mom tensors is protected
67 Note that in dense implementation of this algorithm, ms and mom will
69 and mom will not update in iterations during which the grad is zero.
75 $$mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)$$
76 $$var <- var - mom$$
|
| D | api_def_ResourceApplyRMSProp.pbtxt | 16 name: "mom"
48 If `True`, updating of the var, ms, and mom tensors is protected
55 Note that in dense implementation of this algorithm, ms and mom will
57 and mom will not update in iterations during which the grad is zero.
63 mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)
64 var <- var - mom
|
| D | api_def_ApplyRMSProp.pbtxt | 16 name: "mom"
54 If `True`, updating of the var, ms, and mom tensors is protected
61 Note that in dense implementation of this algorithm, ms and mom will
63 and mom will not update in iterations during which the grad is zero.
69 mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)
70 var <- var - mom
|
| D | api_def_ResourceSparseApplyCenteredRMSProp.pbtxt | 22 name: "mom"
54 A vector of indices into the first dimension of var, ms and mom.
60 If `True`, updating of the var, mg, ms, and mom tensors is
72 Note that in dense implementation of this algorithm, mg, ms, and mom will
74 and mom will not update in iterations during which the grad is zero.
81 mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)
82 var <- var - mom
|
| D | api_def_SparseApplyCenteredRMSProp.pbtxt | 22 name: "mom"
54 A vector of indices into the first dimension of var, ms and mom.
66 If `True`, updating of the var, mg, ms, and mom tensors is
78 Note that in dense implementation of this algorithm, mg, ms, and mom will
80 and mom will not update in iterations during which the grad is zero.
87 $$mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)$$
88 $$var <- var - mom$$
|
| D | api_def_ResourceApplyCenteredRMSProp.pbtxt | 22 name: "mom"
54 If `True`, updating of the var, mg, ms, and mom tensors is
66 Note that in dense implementation of this algorithm, mg, ms, and mom will
68 and mom will not update in iterations during which the grad is zero.
77 mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms - mg * mg + epsilon)
78 var <- var - mom
|
| D | api_def_ApplyCenteredRMSProp.pbtxt | 22 name: "mom"
60 If `True`, updating of the var, mg, ms, and mom tensors is
72 Note that in dense implementation of this algorithm, mg, ms, and mom will
74 and mom will not update in iterations during which the grad is zero.
83 mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms - mg * mg + epsilon)
84 var <- var - mom
|
| D | api_def_RetrieveTPUEmbeddingRMSPropParameters.pbtxt | 17 name: "mom"
19 Parameter mom updated by the RMSProp optimization algorithm.
|
| D | api_def_LoadTPUEmbeddingRMSPropParameters.pbtxt | 17 name: "mom"
19 Value of mom used in the RMSProp optimization algorithm.
|
| D | api_def_RetrieveTPUEmbeddingCenteredRMSPropParameters.pbtxt | 17 name: "mom"
19 Parameter mom updated by the centered RMSProp optimization algorithm.
|
| D | api_def_RetrieveTPUEmbeddingRMSPropParametersGradAccumDebug.pbtxt | 17 name: "mom"
19 Parameter mom updated by the RMSProp optimization algorithm.
|
| D | api_def_LoadTPUEmbeddingRMSPropParametersGradAccumDebug.pbtxt | 17 name: "mom"
19 Value of mom used in the RMSProp optimization algorithm.
|
| D | api_def_LoadTPUEmbeddingCenteredRMSPropParameters.pbtxt | 17 name: "mom"
19 Value of mom used in the centered RMSProp optimization algorithm.
|
| /external/tensorflow/tensorflow/contrib/opt/python/training/ |
| D | lars_optimizer.py | 120 mom = self.get_slot(var, "momentum")
123 mom,
132 mom = self.get_slot(var, "momentum")
135 mom.handle,
144 mom = self.get_slot(var, "momentum")
147 mom,
156 mom = self.get_slot(var, "momentum")
159 mom.handle,
|
| /external/tensorflow/tensorflow/python/keras/optimizer_v2/ |
| D | rmsprop.py | 141 mom = self.get_slot(var, "momentum")
148 mom.handle,
159 mom.handle,
186 mom = self.get_slot(var, "momentum")
193 mom.handle,
205 mom.handle,
|
| D | rmsprop_test.py | 60 def _rmsprop_update_numpy(self, var, g, mg, rms, mom, lr, rho, momentum, argument
70 mom_t = momentum * mom + lr * g / (np.sqrt(denom_t + epsilon))
73 mom_t = mom
77 def _sparse_rmsprop_update_numpy(self, var, gindexs, gvalues, mg, rms, mom, argument
81 mom_t = copy.deepcopy(mom)
93 mom_t[gindex] = momentum * mom[gindex] + lr * gvalue / np.sqrt(denom_t +
97 mom_t[gindex] = mom[gindex]
|
| /external/tensorflow/tensorflow/core/kernels/ |
| D | training_ops_gpu.cu.cc | 230 typename TTypes<T>::Flat ms, typename TTypes<T>::Flat mom, in operator ()()
243 mom.device(d) = in operator ()()
244 mom * momentum.reshape(single).broadcast(bcast) + in operator ()()
247 var.device(d) -= mom; in operator ()()
255 typename TTypes<T>::Flat mom, in operator ()()
270 mom.device(d) = mom * momentum.reshape(single).broadcast(bcast) + in operator ()()
272 var.device(d) -= mom; in operator ()()
|
| D | training_ops_test.cc | 137 auto mom = Scalar(g, 0.01); in Momentum() local
138 test::graph::Multi(g, "ApplyMomentum", {var, accum, lr, grad, mom}); in Momentum()
203 auto mom = Var(g, n); in RMSProp() local
207 test::graph::Assign(g, mom, zero); in RMSProp()
214 auto mom = Var(g, n); in RMSProp() local
221 {var, ms, mom, lr, rho, momentum, epsilon, grad}); in RMSProp()
|